core: further vectorize copyTo with mask

[fengyuentau]

Merge with opencv/opencv_extra#1247.

Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

• I agree to contribute to the project under Apache 2 License.
• To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
• The PR is proposed to the proper branch
• There is a reference to the original bug report and related work
• There is accuracy test, performance test and test data in opencv_extra repository, if applicable
  Patch to opencv_extra has the same branch name.
• The feature is well documented and sample code can be built with the project CMake

[fengyuentau]

Updated: dropped 32SC2 kernel.

Performance results:

i7 12700K, GCC 12.3.0

                    Name of Test                     copyMask-base copyMask-patch copyMask-patch
                                                                                        vs
                                                                                  copyMask-base
                                                                                    (x-factor)
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC1)         0.000         0.000           0.99
Mat_CopyToWithMask::Size_MatType::(127x61, 16UC1)        0.000         0.000           0.99
Mat_CopyToWithMask::Size_MatType::(127x61, 32SC1)        0.001         0.001           0.99
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC2)         0.000         0.000           0.99
Mat_CopyToWithMask::Size_MatType::(127x61, 32SC2)        0.002         0.001           1.46
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC3)         0.001         0.001           0.97
Mat_CopyToWithMask::Size_MatType::(127x61, 16UC3)        0.001         0.001           1.00
Mat_CopyToWithMask::Size_MatType::(127x61, 32FC4)        0.002         0.002           1.00
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC1)      0.102         0.103           0.99
Mat_CopyToWithMask::Size_MatType::(1920x1080, 16UC1)     0.163         0.163           1.00
Mat_CopyToWithMask::Size_MatType::(1920x1080, 32SC1)     0.339         0.340           1.00
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC2)      0.181         0.165           1.10
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC3)      0.237         0.240           0.99
Mat_CopyToWithMask::Size_MatType::(1920x1080, 16UC3)     0.672         0.686           0.98
Mat_CopyToWithMask::Size_MatType::(1920x1080, 32FC4)     3.529         3.527           1.00

M2, Apple Clang 16.0.0

                    Name of Test                     copyMask-base copyMask-patch copyMask-patch
                                                                                        vs
                                                                                  copyMask-base
                                                                                    (x-factor)
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC1)         0.000         0.000           0.98
Mat_CopyToWithMask::Size_MatType::(127x61, 16UC1)        0.001         0.001           0.99
Mat_CopyToWithMask::Size_MatType::(127x61, 32SC1)        0.003         0.001           3.04
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC2)         0.001         0.001           1.04
Mat_CopyToWithMask::Size_MatType::(127x61, 32SC2)        0.003         0.002           1.77
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC3)         0.003         0.001           3.98
Mat_CopyToWithMask::Size_MatType::(127x61, 16UC3)        0.003         0.002           2.08
Mat_CopyToWithMask::Size_MatType::(127x61, 32FC4)        0.003         0.003           0.99
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC1)      0.119         0.116           1.03
Mat_CopyToWithMask::Size_MatType::(1920x1080, 16UC1)     0.151         0.149           1.01
Mat_CopyToWithMask::Size_MatType::(1920x1080, 32SC1)     0.797         0.270           2.95
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC2)      0.156         0.155           1.01
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC3)      0.916         0.226           4.05
Mat_CopyToWithMask::Size_MatType::(1920x1080, 16UC3)     0.974         0.464           2.10
Mat_CopyToWithMask::Size_MatType::(1920x1080, 32FC4)     1.139         1.138           1.00

Orin, GCC 11.4.0

                    Name of Test                     copyMask-base copyMask-patch copyMask-patch
                                                                                        vs
                                                                                  copyMask-base
                                                                                    (x-factor)
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC1)         0.001         0.001           0.98
Mat_CopyToWithMask::Size_MatType::(127x61, 16UC1)        0.001         0.001           0.98
Mat_CopyToWithMask::Size_MatType::(127x61, 32SC1)        0.006         0.003           2.06
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC2)         0.001         0.001           0.99
Mat_CopyToWithMask::Size_MatType::(127x61, 32SC2)        0.006         0.006           0.98
Mat_CopyToWithMask::Size_MatType::(127x61, 8UC3)         0.010         0.004           2.76
Mat_CopyToWithMask::Size_MatType::(127x61, 16UC3)        0.010         0.007           1.39
Mat_CopyToWithMask::Size_MatType::(127x61, 32FC4)        0.010         0.010           1.05
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC1)      0.345         0.329           1.05
Mat_CopyToWithMask::Size_MatType::(1920x1080, 16UC1)     0.783         0.776           1.01
Mat_CopyToWithMask::Size_MatType::(1920x1080, 32SC1)     2.018         1.769           1.14
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC2)      0.779         0.791           0.98
Mat_CopyToWithMask::Size_MatType::(1920x1080, 8UC3)      2.820         1.261           2.24
Mat_CopyToWithMask::Size_MatType::(1920x1080, 16UC3)     3.233         2.734           1.18
Mat_CopyToWithMask::Size_MatType::(1920x1080, 32FC4)     6.006         6.000           1.00

[fengyuentau]

Decided to drop copyMask_<Vec2i> as its performance is not consistent on different platforms.